Heart sounds analysis in joint time-frequency domain

Proposes a new technique of signal processing in joint time-frequency domain as resonance time-frequency analysis for analysis of heart sounds. Generated by mechanical vibration of the heart and cardiovascular system, heart sounds provide abundant physiological information, while the measurement is noninvasive and low cost. The resonance enemy distribution defined here exhibits clearly how signal frequency contents change in time, and can be calculated by a fast algorithm. Applications of this technique to heart sounds analysis show that the frequency contents in the first heart sound S1 and the second heart sound S2 are centred mainly from 100 Hz to 130 Hz while the frequency contents in systolic "kala" murmurs mainly distribute discontinuously from 90 Hz to 210 Hz. This technique is helpful for estimation of functional heart state and diagnosis of heart diseases in clinical research.